§TY2027 current-law ACTC claimant-return forecast
Framing: this is the not-seasonally-adjusted number of claimant returns, not credit dollars or children. It resolves on the first TY2027 IRS SOI Publication 1304 Table 3.3 print. The ledger supplies only an expected 2029-01-01 through 2029-12-31 release window, not an exact IRS calendar day; the registered 2029-12-31 deadline is therefore preserved and this discrepancy is disclosed rather than silently converted into a claimed scheduled date.
Parse official Publication 1304 Table 3.3 workbooks 20in33ar.xls and 21in33ar.xls with the registered irs.actc.total_claims adapter.↳ Official first-print whole-return counts fetched and parsed this run: TY2020 = 19,119,249 (19.119249 million), retrieved 2026-08-04T14:59:09Z; TY2021 = 37,771,612 (37.771612 million), retrieved 2026-08-04T14:59:10Z.Parse official Publication 1304 Table 3.3 workbooks 22in33ar.xls and 23in33ar.xls with the registered irs.actc.total_claims adapter.↳ Official first-print whole-return counts fetched and parsed this run: TY2022 = 18,076,696 (18.076696 million), retrieved 2026-08-04T14:59:15Z; TY2023 = 17,626,084 (17.626084 million), retrieved 2026-08-04T14:59:15Z.Fetch current 26 U.S.C. §24 text from the Office of the Law Revision Counsel.↳ The official code text fetched this run shows §24(d)(1)(B)(i)'s underlying earned-income amount is $3,000 and §24(h)(6) substitutes $2,500 for $3,000.Inspect canonical registration b92e9752beaf38a9e2e735c5066e7c741e29436546e7fab2c8d0568f05355909 and check the catalog slug against specs.json.↳ The public registration fixes period 2027, transform factor 0.000001, condition deadline 2027-12-31, expected release window 2029-01-01 through 2029-12-31, and resolutionDate 2029-12-31. The exact proposed slug appeared 0 times in specs.json when checked this run.Base rate/reference class: the four exact-series first prints for TY2020–TY2023 are 19.119249, 37.771612, 18.076696, and 17.626084 million; these match the registered adapter's verified anchors. Their mean = 23.148410 million, median = 18.597972 million, and range = 17.626084–37.771612 million. TY2021 is a conspicuous policy-regime outlier, but it remains in interval calibration rather than being discarded.
Benchmark and model candidates: latest first-print persistence forecasts 17.626084 million. The thesis_model_candidate_v1 persistence candidate has point/p50 = 17.626084, p10 = 5.121972, p90 = 30.130196, 80% interval = [5.121972, 30.130196], 90% interval = [1.556347, 33.695821], intervalMethod = fallback-prior empirical level dispersion, calibration_n = 4, train cutoff = TY2023, and walk-forward MAE = 12.932630 million across the three available transitions. More elaborate time-series fitting is rejected because four observations with a major TY2021 regime break do not support stable parameter estimation. Persistence is selected.
Prior/update/interval: prior = TY2023 persistence = 17.626084 million, using the four TY2020–TY2023 official first prints. Adjustment components = 0.000000 million for momentum, 0.000000 million for one-offs, and 0.000000 million for policy because the forecast is explicitly conditional on the registered current-law threshold remaining operative and no direct TY2027 filing signal was fetched. Prior weight = 100%; update weight = 0%. For this annual claimant-return flow series, realized level dispersion is sigma = sample_std(19.119249, 37.771612, 18.076696, 17.626084) = 9.768837 million. The 80% half-width is 1.28*sigma = 12.504112 million, so 17.626084 - 12.504112 = 5.121972 and 17.626084 + 12.504112 = 30.130196 million. The interval method is fallback-prior empirical level dispersion; all 4 of 4 historical prints fall within these same-width bounds around the persistence point.
Inside-view restraint: the current-law condition identifies the arm but supplies no direct evidence that TY2027 claimant counts will differ from the latest exact-series print. The dramatic TY2021 value demonstrates policy sensitivity and justifies uncertainty, but using that temporary regime as a directional update under the stated current-law arm would double-count a mechanism not shown to persist.
Stress test: upside risk would land above the interval if an enacted expansion outside this cell's condition, an administrative eligibility change, or an unusually large eligible-filer surge pushed claims above 30.130196 million. Downside risk would land below the interval if labor-income eligibility, take-up, or filing participation collapsed enough to produce fewer than 5.121972 million claimant returns. Evidence that would move the point up is a direct official projection or filing-season count showing broader eligible take-up; evidence that would move it down is a direct IRS count showing sustained claimant attrition.
Review disposition: accepted both optional suggestions by changing the driver to “Latest first-print persistence” and noting that all four historical values match the registered adapter's verified anchors; no required fixes were identified.